JP7218233B2 - Program operation system, program operation method - Google Patents

Description

The present invention relates to a program operating system and a program operating method.

In recent years, with the development of SaaS (Software as a Service) technology, new services/software based on machine/deep learning are being provided as Web services. In particular, if a web service is built by mashing up microservices (programs) provided by a third party, when individual microservices are upgraded due to the addition of new functions or modifications, the addition of microservices , the modified functionality should be tested and the impact of that functionality on the Web service should be evaluated.

Patent document 1 exists as a conventional technique related to testing when upgrading a program. Patent Document 1 aims to provide an efficient technique for upgrading a business system configured from a plurality of modules. In order to achieve this purpose, JP-A-2003-200000 describes managing whether or not each module related to each function to be updated has been changed, and storing the updated function and the changed module in association with each other. specific modules.

Japanese Patent Application Laid-Open No. 2013-228970

In a Web service constructed by mashing up microservices, the microservices may be upgraded to add or change functions. In such a case, it is necessary to test and verify the problems and effects of the upgraded microservice on the entire Web service. In addition, since a plurality of types of Web services are provided and each has a different evaluation axis (KPI: Key Performance Indicator), there is a possibility that whether or not to upgrade the version of the microservice differs depending on the Web service.

In Patent Literature 1, modules related to updated functions are identified and module changes are managed, but evaluation of the impact of the changed modules is not considered. Not disclosed. It also does not consider the case where a module is associated with multiple functions.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a program operating system and a program operating method capable of evaluating the effect of a changed program on a service provided by executing one or more programs.

A program operation system according to an aspect of the present invention includes a pre-update service flow including pre-update microservices and a post-update service flow including post-update microservices in services provided by executing one or more microservices. a microservice management unit for executing a service flow and outputting an execution result of the pre-update service flow and an execution result of the post-update service flow; and an execution result of the pre-update service flow and execution of the post-update service flow. a service update analysis processing unit that evaluates the results and selects either the pre-update service flow or the post-update service flow as a service flow in which the evaluation result satisfies a predetermined relationship. It is configured as a program operation system that

According to one aspect of the present invention, the impact of a modified program on a service provided by executing one or more programs can be evaluated.

Diagram showing the system configuration of this embodiment A diagram showing an outline of the target processing of this embodiment. Microservice management table of this embodiment Service flow evaluation processing management table of this embodiment Service flow management table of this embodiment Web service management table of this embodiment Flowchart of this embodiment Service flow management table after updating in this embodiment Web service management table after updating in this embodiment Microservice management table after updating in this embodiment

Embodiments of the present invention will be described below with reference to the drawings. The following description and drawings are examples for explaining the present invention, and are appropriately omitted and simplified for clarity of explanation. The present invention can also be implemented in various other forms. Unless otherwise specified, each component may be singular or plural.

The position, size, shape, range, etc. of each component shown in the drawings may not represent the actual position, size, shape, range, etc., in order to facilitate understanding of the invention. As such, the present invention is not necessarily limited to the locations, sizes, shapes, extents, etc., disclosed in the drawings.

In the following description, various types of information may be described using expressions such as “table” and “list”, but various types of information may be expressed in data structures other than these. "XX table", "XX list", etc. are sometimes referred to as "XX information" to indicate that they do not depend on the data structure. When describing identification information, expressions such as “identification information”, “identifier”, “name”, “ID”, and “number” are used, but these can be replaced with each other.

When there are a plurality of components having the same or similar functions, they may be described with the same reference numerals and different suffixes. However, if there is no need to distinguish between these multiple constituent elements, the subscripts may be omitted in the description.

Also, in the following description, there are cases where processing performed by executing a program is described, but the program is executed by a processor (for example, CPU, GPU) to appropriately perform the specified processing using storage resources ( For example, a memory) and/or an interface device (for example, a communication port) or the like is used, so processing may be performed by a processor. Similarly, a main body of processing executed by executing a program may be a controller having a processor, a device, a system, a computer, or a node. The subject of the processing performed by executing the program may be an arithmetic unit, and may include a dedicated circuit (for example, FPGA or ASIC) that performs specific processing.

A program may be installed on a device, such as a computer, from a program source. The program source may be, for example, a program distribution server or a computer-readable storage medium. When the program source is a program distribution server, the program distribution server may include a processor and storage resources for storing the distribution target program, and the processor of the program distribution server may distribute the distribution target program to other computers. Also, in the following description, two or more programs may be implemented as one program, and one program may be implemented as two or more programs.

FIG. 1 is a diagram showing the configuration of the program operating system of this embodiment. As shown in FIG. 1, the program operation system 1000 includes a microservice management server 10, a service flow management server 11, a web service operation server 12, and a microservice operation server 13, which are connected via a network N. Connected. Application A 141 , application B 142 , and application C 143 are applications installed on the computer of the user who uses the web service executed by the web service execution unit 120 .

The servers and computers that this system has include, as hardware, storage devices such as HDD (Hard Disk Drive) and SSD (Solid State Drive), CPU (Central Processing Unit), RAM (Random Access Memory), ROM (Read It has a configuration of a general information processing device having an arithmetic device with a memory only. Also, the functions of the functional units of each server are implemented by executing a program. For example, the functions of each section are realized by the CPU reading out and executing a program from the ROM. The program may be provided to each server by being read from a storage medium such as a USB (Universal Serial Bus) memory or downloaded from another computer via a network.

The Web service is a service that operates according to a service flow managed by the service flow management unit 110. FIG. For example, it corresponds to the processing flow of a program defined by connecting nodes representing so-called program processing by a visual programming tool such as Node-RED. If the processing flow is Node-RED, Flow. It is defined in a file format described in a predetermined format such as json, and by referring to, analyzing, and processing this file, the processing shown below can be realized. Also, the processing flow of a program defined in Node-RED becomes available as a Web service that actually operates by being deployed on Node-RED. Also, microservices are implemented as containers of Docker (registered trademark), which is a virtualization environment, for example.

Therefore, for example, the microservice management server 10 is a server that provides a function of managing the Docker container. The service flow management server 11 is a server that provides a function of managing the processing flow. The Web service operating server 12 is a server on which Node-RED operates. The microservice running server 14 can be implemented as a server running Docker.

A service flow in this embodiment includes processing of one or more microservices, and the microservices call, for example, a service running in the microservice execution unit 150 in the flow.

The microservices are managed by the microservice management unit 100 of the microservice management server 10 . Then, when an update such as a version upgrade of the microservice occurs, the microservice management unit 100 temporarily generates a service flow reflecting the update (updated service flow), and parallelizes it with the same input as the service flow before the update. Generate a working flow. Furthermore, the microservice management unit 100 deploys a Web service to which a flow for comparing and evaluating the output of the updated service flow and the output of the service flow before updating is added.

Then, every time the Web service is used from the application, the updated service flow and the service flow before update can be evaluated. As a result of the evaluation, either the update service flow or the pre-update flow is selected. The selected service flow is newly deployed as a web service, and the service flow management table 400 and microservice management table 200 are updated. As a result of updating the macroservice management table 200, microservices that are old version microservices that do not have any related services are deleted.

The details of this process will be described below with reference to the drawings. FIG. 2 is a diagram showing an overview of target processing. Services operating in the microservice execution unit 150 are managed by the macroservice management table 200 of the microservice management unit 100 . In this system, the microservice management unit 100 executes one or a plurality of microservices (programs) to provide pre-update service flows including pre-update microservices. , and the post-update service flow including the post-update microservice, output the execution result of the pre-update service flow and the post-update service flow, and the service update analysis processing unit 131 of the service evaluation update processing unit 130 evaluates the execution result of the pre-update service flow and the execution result of the post-update service flow, and selects either the pre-update service flow or the post-update service flow as the service flow for which the evaluation result satisfies a predetermined relationship. and registered as a new service flow. This makes it possible to evaluate the impact of a changed program on services provided by executing one or more programs. For example, even if a change occurs in the program, It is possible to automatically evaluate which service is suitable as a service to be provided.

Further, since the service update analysis processing unit 131 selects a service flow with a high evaluation result as a service flow whose evaluation result satisfies a predetermined relationship, the service flow with a high evaluation (for example, KPI value) is executed. By doing so, it is possible to provide a service with a higher degree of satisfaction than the service before the update. Below, a service flow with a high KPI value is described as an example of a service flow that satisfies a predetermined relationship. It may be determined whether or not the service flow before and after the update satisfies a predetermined relationship.

FIG. 3 shows a configuration diagram of the microservice management table 200. As shown in FIG. As shown in FIG. 3, the macro service management table 200 includes a micro service name 201, a program version 202 indicating the program version, a model version 203 indicating program learning data, parameter sets, etc., and the micro service being used. It consists of related services 204 that show the service flow. For example, microservice MC_A in the figure indicates that program version 1.0, no model version, and related services are service A, service B, and service A1.

The service flow management unit 110 manages the service flow constructed by combining the microservices using the service flow management table 400 .

FIG. 5 shows a configuration diagram of the service flow management table 400. As shown in FIG. As shown in the figure, it consists of a service flow name 401 , a service flow version 402 and flow information 403 . For example, service SV_A (service A) is version 1.0, and the flow is configured to operate in order of microservices MC_A, MC_B, and MC_C.

A service flow managed by the service flow management unit 110 can operate as a web service that can be used from an application by being deployed to the web service execution unit 120 .

Each service flow defines a process for evaluating the result of the process, and the process is managed in the service flow evaluation process management table 300 of the service evaluation update processor 130 .

FIG. 4 shows a configuration diagram of the service flow evaluation processing management table 300. As shown in FIG. As shown in the figure, it consists of a service flow name 301 , a KPI 302 that is an evaluation process index, and an evaluation process name 303 . For example, service SV_A is a process of evaluating by detection rate, and indicates evaluation by evaluation process EV_A1. It is also possible to prepare a plurality of evaluation processes in a service flow, and an evaluation process is selected when deployed as a web service.
A deployed web service is managed by the web service management table 500 .

FIG. 6 shows a configuration diagram of the Web service management table 500. As shown in FIG. As shown in the figure, the web service management table 500 includes a web service name 501 , service flow name 401 , service flow version 402 and evaluation processing 303 . For example, the service A of the Web service is the service flow SV_A of version 1.0 deployed, and the evaluation process EV_A1 is executed.

In the microservice execution unit 150 of FIG. 2, MC_A, MC_B, MC_C, and MC_D are in operation, and an updated version MC_B1 of the MC_B microservice is newly provided.

Details of the processing of this embodiment will be described below with reference to the flowchart of FIG. 7 .
In FIG. 7, it is assumed that all services including updated microservices are executed.

The microservice manager 100 detects execution of microservice MC_B1, which is an updated version of microservice MC_B. As a detection method, for example, an updated microservice updated by a microservice developer is registered in the microservice management table 200 by the microservice management unit 100, or an update is detected in cooperation with a version control system such as Git. (Step 1).

Next, the microservice management unit 100 registers the detected microservice in the microservice management table 200 . In the example of FIG. 3, the microservice management unit 100 detects an update of MC_B and registers an updated version of MC_B1, and the updated program version and model version are set (Step 2).

Next, the microservice management unit 100 checks the existence and configuration of related web services of the microservice MC_B whose update has been detected, from the microservice management table 200 . In the example of FIG. 3, the microservice management unit 100 confirms that service A, service B, and service C exist as related Web services of MC_B (Step 3).

The microservice management unit 100 processes service flow information corresponding to the service flow and version of each Web service. If there is an unprocessed related Web service (Step 4; Yes), the process proceeds to Step 5, and if there is no unprocessed related Web service (Step 4; No), the process ends (Step 4).

In the following, a case where service A is an unprocessed related web service will be described as an example. The microservice management unit 100 refers to the web service management table 500 shown in FIG. 6, and sets SV_A, Ver. as 1.0. The microservice management unit 100 refers to the service flow management table 400 , generates an updated service flow in which MC_B is changed to MC_B1, and registers the updated service flow in the service flow management table 400 . In FIG. 5, the microservice management unit 100 registers the service flow SV_A1 as the updated flow of SV_A. The updated flow information becomes MC_A→MC_B1→MC_C (Step 5).

Next, the microservice management unit 100 generates a web service of the service flow generated in step 5 and registers the web service in the web service management table 500. FIG. In the case of SV_A1, the microservice management unit 100 registers service A1 in the web service management table 500 of FIG. 6 as an updated web service. The microservice management unit 100 registers EV_A1, which is the same as service A before update, as an evaluation process, and registers SV_A1 as a service flow. The microservice management unit 100 also updates the related web service list of the microservice management table 200 when a web service is generated. In FIG. 3, the microservice management unit 100 adds service A1, which is the updated web service, to microservices MC_A, MC_B1, and MC_C (Step 6).

Next, the microservice management unit 100 adds a process of comparing and evaluating the updated web service generated in step 6 and the original web service. When the services to be compared and evaluated are service A1 and service A, the microservice management unit 100 inputs the result of each service to the evaluation process EV_A1. The evaluation process includes a process of evaluating service results using a predetermined KPI, a process of comparing and analyzing the evaluation results of the respective services, and a process of recording the analysis results in a log (Step 7).

The application can continue to use the original Web service even in a situation where the comparison evaluation process of the updated Web service and the original Web service is deployed in Step 6. Service A) and the updated Web service (for example, service A1) are processed in parallel, and the execution results of both and the comparison results of both are stored in a log. The service update analysis processing unit 131 evaluates which service is superior by analyzing the log after a predetermined number of times of use or after a predetermined period of time has elapsed, and as a result, determines whether or not the Web service can be updated. (Step 8).

In the log analysis, the microservice management unit 100 executes the pre-update service flow and the post-update service flow a predetermined number of times. Using the statistical value of the evaluation of the execution result of the pre-update service flow and the statistical value of the evaluation of the execution result of the service flow after update, a service flow with a high evaluation result is selected and registered as a new service flow. You can do it. Various statistical values such as average values and median values can be used as the statistical values.

If the service update analysis processing unit 131 determines that the original web service is superior as a result of the evaluation and that the web service should not be updated (Step 9; No), it executes update cancellation processing (Step 10). do. On the other hand, when the service update analysis processing unit 131 determines that the original web service is not good and that the web service may be updated (Step 9; Yes), it proceeds to update processing (Step 11) (Step 9).

In the update cancellation process, the service update analysis processing unit 131 deletes the updated Web service and the comparison evaluation process. At the same time, the corresponding Web service in the Web service management table 500 and the corresponding service flow in the service flow management table 400 are deleted. Also, the service update analysis processing unit 131 deletes the relevant web service from the related web service list of the microservice management table 200 . In FIG. 3, the service update analysis processing unit 131 deletes service A1 from the associated web service list of MC_A, MC_B1, and MC_C (Step 10).

In the update process, the service update analysis processing unit 131 deletes the original Web service and the comparative evaluation process. The service update analysis processing unit 131 also temporarily deletes service A1 of the updated web service, deletes service A from the web service management table 500, and changes service A1 to service A. FIG. Further, the service update analysis processing unit 131 changes SV_A1 in the service flow management table 400 to SV_A, and after registering the version obtained by incrementing the version of the original service flow, changes the version 1.1 of the service flow SV_A to Web service. Deploy as A.

FIG. 8 shows the service flow management table after update processing. FIG. 9 shows the Web service management table after update processing. In FIG. 8, it can be seen that the service update analysis processing unit 131 registers the updated service flow SV_A1 as version 1.1 of SV_A. Further, in FIG. 9, it can be seen that the service update analysis processing unit 131 deletes the updated Web service A1 and registers the Web service A as version 1.1. Further, the service update analysis processing unit 131 deletes service A from the related web service list of the microservice management table 200 and changes service A1 to service A. FIG. FIG. 10 shows the microservice management table after update processing. In FIG. 10, as a result of deleting service A and changing service A1 to service A, for example, only service A and service C are registered in the associated web service list of microservice MC_C. (Step 11).

After completing Step 10 or Step 11, the microservice management unit 100 updates the microservice management table 200. FIG. In the macro service management table 200, the micro service management unit 100 deletes the record of the micro service if it is an old version of the micro service and the related web service list is blank, that is, if the web service being used does not exist. (Step 12).

In this way, the service update analysis processing unit 131 selects the evaluated service (eg, SV_A) among the plurality of services including the microservice (eg, MC_B) included in the selected service flow (eg, SV_A). other services than SV_B (e.g. SV_B) such that the modified microservice (e.g. MC_B) is no longer selected from all other services (e.g. service flows containing MC_B such as SV_B) ( That is, MC_B before the change is no longer highly evaluated), the service flow for which the selection of the service flow before update or the service flow after update is not performed (for example, the service evaluated with a low KPI value) flow) is deleted from the program operating system 1000 . Therefore, when changing a microservice included in a certain service, it is possible to eliminate erroneous deletion of the changed microservice used in other services or a service including the microservice.

As described above, the processing of one Web service is completed by the processing from Step 5 to Step 12 . Next, the process moves to Step 4 again and repeats the process until there are no more unprocessed Web services.

As described above, in this embodiment, the microservice management unit 100 analyzes the configuration information of the web service to determine whether or not there is a related web service for each version of the microservice, and records the result as a table. , the service flow management unit 110 manages service flow configuration information representing service flow information indicating the flow of processing of microservices that constitute the Web service, and the Web service execution unit 120 manages the service flow configuration information. In a system where a service can be deployed and run based on , when a version update of a microservice occurs, all running web services related to other versions of that microservice will be updated to the updated microservice. Configure and deploy a superseded update service flow. In this system, a KPI for evaluating the Web service is calculated, and a process for comparing the original Web service and the updated Web service is deployed. Then, when a service request is received from a user via a Web service, this system executes processing of both the original Web service and the update Web service for the request, and the result of both services is KPI is calculated from , and the comparison result is recorded in the log.

Then, after the service requests have been executed a predetermined number of times, the accumulated logs are analyzed, and if the KPI has not deteriorated, the original Web service and the updated Web service are replaced, and a Web service configuration information table and Update service flow configuration information. On the other hand, if the KPI is deteriorating, delete the update web service. As a result, for example, when a business service provided as a web service is built with a mashup of multiple subprograms, so-called microservices, and the service is updated, when the version of the microservice is updated, It becomes possible to manage without degrading the quality of Web services that utilize the microservices.

REFERENCE SIGNS LIST 100 Microservice management unit 200 Microservice management table 110 Service flow management unit 120 Web service execution unit 130 Service evaluation update processing unit 141 to 143 Applications 150 Microservice execution unit 300 ... Service evaluation processing management table 400 ... Service flow management table 500 ... Web service management table

Claims (6)

In a service provided by executing one or more microservices, executing a pre-update service flow including pre-update microservices and a post-update service flow including post-update microservices, and said pre-update service flow a microservice management unit that outputs the execution result of and the execution result of the updated service flow;
Either the pre-update service flow or the post-update service flow is a service flow that evaluates the execution result of the pre-update service flow and the post-update service flow, and the evaluation result satisfies a predetermined relationship. a service update analysis processing unit that selects the
The service update analysis processing unit evaluates a service other than the evaluated service among a plurality of services including the microservice included in the selected service flow, and the microservice performs the evaluation. deleting from the program operation system the service flow that was not selected from the pre-update service flow or the post-update service flow when it is no longer selected from other services;
A program operation system characterized by: The program operating system according to claim 1,
The service update analysis processing unit selects a service flow with a high evaluation result as a service flow whose evaluation result satisfies a predetermined relationship.
A program operation system characterized by: The program operating system according to claim 2,
The microservice management unit executes the pre-update service flow and the post-update service flow a predetermined number of times,
The service update analysis processing unit uses the statistical value of the evaluation of the execution result of the pre-update service flow obtained by executing the predetermined number of times and the statistical value of the evaluation of the execution result of the post-update service flow, selecting a service flow with a high evaluation result;
A program operation system characterized by: A microservice management unit executes a pre-update service flow including pre-update microservices and a post-update service flow including post-update microservices in services provided by executing one or more microservices. ,
the microservice management unit outputs an execution result of the pre-update service flow and an execution result of the post-update service flow;
a service update analysis processing unit evaluating an execution result of the pre-update service flow and an execution result of the post-update service flow;
When the service update analysis processing unit selects either the pre-update service flow or the post-update service flow as a service flow whose evaluation result satisfies a predetermined relationship ,
The service update analysis processing unit evaluates a service other than the evaluated service among a plurality of services including the microservice included in the selected service flow, and the microservice performs the evaluation. deleting from the program operation system the service flow that was not selected from the pre-update service flow or the post-update service flow when it is no longer selected from other services;
A program operation method characterized by: The program operating method according to claim 4 ,
The service update analysis processing unit selects a service flow with a high evaluation result as a service flow whose evaluation result satisfies a predetermined relationship.
A program operation method characterized by: The program operating method according to claim 5 ,
The microservice management unit executes the pre-update service flow and the post-update service flow a predetermined number of times,
The service update analysis processing unit uses the statistical value of the evaluation of the execution result of the pre-update service flow obtained by executing the predetermined number of times and the statistical value of the evaluation of the execution result of the post-update service flow, selecting a service flow with a high evaluation result;
A program operation method characterized by:

Images